Does Bing Webmaster Tools Show AI Visibility Data for Microsoft AI Products?

What does the AI Performance report measure?

From Microsoft's launch post and help documentation, the report includes:

The report covers Microsoft Copilot, AI-generated summaries in Bing, and what Microsoft calls select partner integrations, which it does not name.

Does Bing Webmaster Tools’ AI Performance report show actual citation data?

Yes, with one caveat. The counts come from Microsoft's own logs of when your pages were shown as sources in its AI answers, not from prompts a tracker chose to run, but Microsoft says the data shown "represents a sample of overall citation activity". Treat the numbers as a consistent first-party sample: good for trends and comparisons, not exact totals. That is its main advantage over third-party trackers, which run their own questions and extrapolate. The limits are what it leaves out. There is no click data, so you cannot see which citations sent visitors. There is no answer text, so you cannot see how your brand was described. And it covers only Microsoft's surfaces. Citation counts also reflect how often pages were shown, not whether they were the main source or one of many.

How does Bing Webmaster Tools provide Copilot grounding data?

Through grounding queries. When Copilot or Bing's AI answers a question, it searches the web with one or more short queries, then uses the results to write the answer. Microsoft calls these grounding queries, and the report lists the ones that led to your pages being cited. They are shorter and plainer than user prompts: a user asking a long question about choosing payroll software for a construction firm might produce grounding queries such as "payroll software construction" or "construction payroll compliance."

That makes them useful in two ways. They show the language Microsoft's AI uses to retrieve content, which you can match in page titles, headings and opening sentences. And they give you a ready-made list of questions to track in other assistants. Microsoft's June update grouped them by intent and topic, so a long list becomes easier to read. How Copilot's short Bing queries work is explained further in why Bing matters for AI search visibility.

How to find which Bing AI grounding queries map to each page

Use the Grounding Query – Pages mapping in the report, which rolled out from March 2026:

1. Open AI Performance in Bing Webmaster Tools for your verified site, and choose a date range.
2. Select a grounding query to see the pages cited for it, or select a page to see the grounding queries it was cited for.
3. Note that one query can map to several pages, and one page to several queries.
4. Export the filtered view to CSV or Excel. Microsoft's help page says you can filter by a query or a page, but not both at once, so export and cross-reference when you need both.
5. Expect small differences in counts depending on which direction you filter from, as reviewers have noted.

With the export, build a sheet of each important page, its grounding queries, their intents and your citation share. That shows which pages win which questions, and where a page is cited for queries you did not intend.

What can you do with the data?

• Validate your best pages. The pages cited most are your AI reference pages; keep them current and accurate.
• Find gaps. Pages that are indexed but rarely cited, for queries you care about, are candidates for clearer answers and structure.
• Fix mismatches. If a page is cited for queries it does not answer well, tighten it or create a better page for those queries.
• Track share, not just counts. Citation share for a grounding query shows whether you are gaining ground on it even when total citations move with demand.
• Use grounding queries elsewhere. Add them to the prompt list you check in ChatGPT, Perplexity and Google.
• Submit updates quickly. Microsoft recommends IndexNow, so AI answers use the latest version of a page.

CXL, analysing its own report, found its ten most-cited pages ranked around fifth on average in organic search. That fits Microsoft's model: AI answers are grounded in search results, so pages that rank well in Bing tend to be cited.

Why do the numbers change?

Microsoft's documentation lists several reasons citation patterns move: changes in user demand, model updates, partner refresh cycles, freshness, and changes elsewhere on the web. Treat a sudden change as a prompt to investigate, not as a verdict on your content. Compare periods with the Compare feature, check whether the change is concentrated in one topic or intent, and look at the pages involved before rewriting anything.

What are the report's limits?

• No clicks or referrals. Pair it with analytics, such as GA4's AI Assistant channel or Microsoft Clarity's AI Visibility dashboard, which reports AI referral traffic.
• No answer text or prompts. Grounding queries are the AI's search terms, not what users typed.
• No competitors. Citation share gives your proportion but not who else was cited.
• Microsoft surfaces, plus unnamed partners. Do not read it as data for ChatGPT, Gemini, Claude or Perplexity.
• No API announced. Microsoft did not announce one with the preview, so plan on manual exports.
• Preview labels. Intents and topics come from classifiers Microsoft describes as still evolving.

Joining citations to traffic is covered in how to correlate AI visibility metrics with website traffic, referral setup in how to track ChatGPT and other AI referrals in GA4, and Clarity in whether Microsoft Clarity's AI visibility dashboard improves GEO.

How should you set it up?

1. Verify your site in Bing Webmaster Tools, or import it from Google Search Console.
2. Submit your sitemap and turn on IndexNow.
3. Wait for data; small sites may see few citations at first.
4. Export monthly: citations, cited pages and the query-to-page mapping.
5. Review by topic and intent, and compare with the previous period.

What might this look like in practice?

Consider an illustrative HR software company, Rotaline. Its AI Performance report showed 1,400 citations in a month, but the query-to-page mapping revealed that most came from one old blog post, cited for grounding queries about holiday entitlement calculations. Its product pages were rarely cited, and its citation share for "shift scheduling software" queries was low. Rotaline updated the blog post with current rules, added a calculator link to its scheduling product, rewrote the product page's opening around the questions in the grounding queries, and submitted both via IndexNow. The next month's Compare view showed product page citations rising, and GA4 showed more Copilot referrals to them.

What do people ask about Bing's AI Performance report?

Is the AI Performance report free?

Yes. It is part of Bing Webmaster Tools, for verified sites.

Does it include ChatGPT?

Not as a named source. It covers Microsoft Copilot, Bing's AI-generated summaries and select partner integrations that Microsoft does not name, so track ChatGPT separately.

Can I see which pages each grounding query cited?

Yes. The query-to-page mapping lets you select a query to see its pages, or a page to see its queries.

Does it show clicks from Copilot?

No. Use analytics to measure visits from Copilot.